1. Field of the Invention
The present invention relates to an image processing device for displaying an image of the surroundings of a vehicle captured by at least one image capturing camera mounted on the vehicle, and particularly to a technique of converting captured images into an image looking down from a point of view above the vehicle and displaying the converted image.
2. Description of the Related Art
Along with an increase in the use of image capturing cameras mounted on automobiles, applications using captured image data have been increasing. Many of such applications are for assisting a driver in driving and for improving the safety of a vehicle in driving and parking of the vehicle. For example, in a system which captures an image of an obstacle present near a vehicle and displays the captured image, images of the surroundings of the vehicle are captured by a plurality of super wide-angle image capturing cameras, and image data of the captured images are synthesized to display a 360° image of the surroundings of the vehicle. In this process, the data of the captured images can also be converted into an image looking down from a point of view above the vehicle (hereinafter referred to as a top view image) and can be displayed on a display.
Japanese Unexamined Patent Application Publication No. 2006-27556 discloses a technique of changing a method of outputting an image of the surroundings of a vehicle in accordance with the presence or absence of an obstacle near the vehicle. When the gear shift position is in reverse, an overhead image is displayed. In this process, a distance sensor determines whether or not an obstacle is present near the vehicle. If the presence of an obstacle is determined, the display is switched to a captured image directed in a direction showing the obstacle.
Japanese Unexamined Patent Application Publication No. 2002-19556 relates to a monitoring system for monitoring the surroundings of a vehicle. In the generation of a top view image, the image of an object becomes unclear in areas in which the ranges of the fields of vision of a plurality of cameras overlap. Therefore, images showing the overlapping areas are not synthesized but are individually displayed in an exception mode.
FIG. 1 is a conceptual diagram for explaining a method of generating a top view image performed by an image processing device. An image capturing camera 12 installed at a side-view mirror of a vehicle 10 captures an image of the left side of the vehicle 10, and outputs data of the captured image. A top view image displaying the surroundings of the vehicle is generated on the basis of the ground of the surroundings of the vehicle shown by the data of the captured image, and shows a display area 14. If the display area 14 includes a columnar solid object 16, for example, the solid object 16 is shown in the top view image as a solid object display 18 extending from the direction of the image capturing camera 12 to the direction of the ground and having a widened upper portion. Further, because the top view image is generated on the basis of the ground of the surroundings of the vehicle, the top view image cannot always accurately display the entire image of the solid object 16. Therefore, the solid object 16 may be displayed with an upper portion thereof widened, or the upper portion of the solid object 16 may not be displayed.
FIGS. 2A and 2B are diagrams illustrating the relationship between a side view image and the top view image. The side view image is an image captured by an image capturing camera installed at a side-view mirror as illustrated in FIG. 1. When the top view image is generated by performing viewpoint conversion on data of captured images of the front side, the rear side, and the lateral sides of a vehicle, a solid object disappears or is displayed in a distorted form in joined portions of the respective images. Further, for example, a side view image 20 of FIG. 2A which captures the image of a lateral side of the vehicle clearly shows a human FIG. 22. Meanwhile, a top view image 24 of FIG. 2B displays only a part of the human FIG. 22 due to the range of a screen to be displayed and the influence of a joined portion 26. Furthermore, although a human FIG. 23 appearing in FIG. 2A is shown in an edge region of the top view image 24, the human FIG. 23 may be displayed with an upper portion thereof widened or the upper portion of the solid object may not be displayed in the top view image 24 of FIG. 2B due to the limitation of the display range of the image. As a result, a user cannot accurately identify from the top view image the solid object present near the vehicle. Thus, it is difficult to use the top view image for parking assistance and so forth.